comparative genomics of the mating-type loci of the mushroom...

Comparative Genomics of the Mating-Type Loci of theMushroom Flammulina velutipes Reveals WidespreadSynteny and Recent InversionsArend F. van Peer1, Soon-Young Park1, Pyung-Gyun Shin1, Kab-Yeul Jang1, Young-Bok Yoo1, Young-Jin

Park2, Byoung-Moo Lee2, Gi-Ho Sung1, Timothy Y. James3, Won-Sik Kong1*

1 Mushroom Research Division, National Institute of Horticultural and Herbal Science, Rural Development Administration, Suwon, Republic of Korea, 2 National Academy

of Agricultural Science, Rural Development Administration, Suwon, Republic of Korea, 3 Department of Ecology and Evolutionary Biology, University of Michigan, Ann

Arbor, Michigan, United States of America

Abstract

Background: Mating-type loci of mushroom fungi contain master regulatory genes that control recognition betweencompatible nuclei, maintenance of compatible nuclei as heterokaryons, and fruiting body development. Regions nearmating-type loci in fungi often show adapted recombination, facilitating the generation of novel mating types and reducingthe production of self-compatible mating types. Compared to other fungi, mushroom fungi have complex mating-typesystems, showing both loci with redundant function (subloci) and subloci with many alleles. The genomic organization ofmating-type loci has been solved in very few mushroom species, which complicates proper interpretation of mating-typeevolution and use of those genes in breeding programs.

Methodology/Principal Findings: We report a complete genetic structure of the mating-type loci from the tetrapolar,edible mushroom Flammulina velutipes mating type A3B3. Two matB3 subloci, matB3a that contains a unique pheromoneand matB3b, were mapped 177 Kb apart on scaffold 1. The matA locus of F. velutipes contains three homeodomain genesdistributed over 73 Kb distant matA3a and matA3b subloci. The conserved matA region in Agaricales approaches 350 Kband contains conserved recombination hotspots showing major rearrangements in F. velutipes and Schizophyllum commune.Important evolutionary differences were indicated; separation of the matA subloci in F. velutipes was diverged from theCoprinopsis cinerea arrangement via two large inversions whereas separation in S. commune emerged through transpositionof gene clusters.

Conclusions/Significance: In our study we determined that the Agaricales have very large scale synteny at matA (,350 Kb)and that this synteny is maintained even when parts of this region are separated through chromosomal rearrangements.Four conserved recombination hotspots allow reshuffling of large fragments of this region. Next to this, it was revealed thatlarge distance subloci can exist in matB as well. Finally, the genes that were linked to specific mating types will serve asmolecular markers in breeding.

Citation: van Peer AF, Park S-Y, Shin P-G, Jang K-Y, Yoo Y-B, et al. (2011) Comparative Genomics of the Mating-Type Loci of the Mushroom Flammulina velutipesReveals Widespread Synteny and Recent Inversions. PLoS ONE 6(7): e22249. doi:10.1371/journal.pone.0022249

Editor: Kirsten Nielsen, University of Minnesota, United States of America

Received March 24, 2011; Accepted June 17, 2011; Published July 20, 2011

Copyright: � 2011 van Peer et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Funding: This work was supported by a grant from the Next-Generation BioGreen 21 Program (No. PJ008216), Rural Development Administration, Republic ofKorea. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The present study was initiatedby the first author who is an employee of the Mushroom Research Division. The Mushroom Research Division is a non-profit organization by the governmentalRural Development Administration in Korea. Research and findings are supportive to this organization for improvement of breeding technologies.

Competing Interests: The authors have declared that no competing interests exist.

* E-mail: [email protected]

Introduction

The genes that regulate mating in fungi encode sets of

regulatory and signaling molecules that are broadly distributed

in eukaryotes. The corresponding pathways, which are often

linked to pathogenicity and fruiting body formation, comparably

control developmental processes in other organisms, such as

pattern formation in development and sexual differentiation in

animals. Studies in Agaricomycetes, that include many important

mushroom-forming fungi, therefore also inform our collective

understanding of cellular development in eukaryotes. Mushroom

forming fungi are further important sources for food (mushrooms,

fermentation), medicine (pathogens, fibers, health-promoting, anti-

cancer products) and green technologies (waste recycling,

fertilizers, bioremediation). As such they represent a massive

economical asset and better understanding of their sexual

propagation is desirable.

Mating is the beginning step in sexual development of

mushroom-forming fungi. Their life cycle is characterized by

haploid as well as diploid stages. Sexual, haploid spores that are

dispersed by a mushroom develop into monokaryotic mycelia.

Upon fusion of two genetically different monokaryons a dikaryotic

mycelium is established and the different nuclei of the two mating

partners coexist within the cells of the dikaryon during vegetative

PLoS ONE | www.plosone.org 1 July 2011 | Volume 6 | Issue 7 | e22249

growth and production of fruiting bodies. Only upon maturation

of the mushroom, the different nuclei fuse in specialized

reproductive cells termed basidia and new haploid spores with

separate mating-types are created (for a review see [1]). The

heterothallic fungi restrict their self-mating by use of one (bipolar

species) or two (tetrapolar species) incompatibility loci. This system

supports out-breeding and helps to promote genetic variability in

populations.

Our current knowledge on the molecular genetics of mating in

mushrooms is primarily based on studies in the model organisms

Schizophyllum commune and Coprinopsis cinerea ( = Coprinus cinereus). The

mating-type loci in those species are termed A and B and control

different developmental pathways that are required to maintain a

fertile dikaryon (for reviews see [2,3]). Each mating-type locus

consists of tightly linked subloci and encodes multiallelic genes

[4,5,6]. These alleles are highly polymorphic at DNA and amino

acid (AA) levels [7,8,9] due to balancing selection that favors

mating-type alleles that become rare and thus extends the

coalescence time between alleles [10]. The A and B subloci are

functionally redundant and heterozygosity at a single sublocus is

sufficient to activate the respective A or B pathway.

A mating-type loci encode pairs of divergently transcribed

homeodomain genes (HD genes) and are typically accompanied by

the Mitochondrial Intermediate Peptidase gene (MIP) and a Beta-flanking

gene [11,12]. HD proteins are distinguished based on conserved

DNA binding motifs; homeodomain 1 (HD1) and homeodomain 2

(HD2). HD1 proteins further contain two nuclear localization

signals, an activation domain and only weakly bind DNA [13,14]

whereas HD2 proteins lack these domains but have strong DNA

binding properties [15,16]. Both HD1 and HD2 proteins possess

N-terminal dimerization motifs that facilitate their interaction.

Interaction of HD1 with a compatible HD2 protein generates a

heterodimer that serves as a transcription factor for the A pathway

[15,17].

The B mating-type loci are comprised of pheromone receptors

and pheromones (recently reviewed in [3]) and each pheromone

receptor is accompanied by one or several pheromones in a

sublocus [9,18]. Pheromone genes encode small precursor proteins

with C-terminal CAAX motifs (C = cysteine, A = aliphatic and X

is any residue) that are farnesylated. They usually contain an acidic

AA pair as well (ER or EH in C. cinerea) about 10–15 AA from the

C-terminus. Those acidic residues are highly conserved in C. cinerea

and have been speculated to be the site of proteolytic cleavage

[5,19,20,21,22]. They are also conserved in other basidiomycete

pheromones, though not in all [23,24]. After farnesylation and

proteolytic splicing, fungal pheromones typically constitute about

9–11 amino acid peptides [19,25].

Fungal pheromone receptors are classified within the Rhodop-

sin-like superfamily and typically contain seven membrane

spanning regions (7-TM). They are further characterized by a

short N-terminal extracellular domain and a long cytoplasmic C-

terminal tail [26]. The cytoplasmic domains are presumed to dock

trimeric G-proteins that can activate the downstream B pathway

after phosphorylation [26]. Phosphorylation of G-proteins is

triggered by interaction of specific pheromones with extracellular

pheromone receptor domains.

The fundamental composition of the A and B mating-type loci

was found to be strongly conserved within basidiomycetes. The

redundant subloci are a result of doubling during evolution and

tetrapolar mating systems were found in all three major linages,

suggesting an ancient origin [27,28]. Moreover, bipolar and

tetrapolar species have been shown to contain essentially the same

genes [24,29]. On the other hand, the new availability of genome

sequences reveals significant variability within mating-type loci of

basidiomycetes and their numbers of pheromone receptor,

pheromone and homeodomain encoding genes differ greatly

between species [12,30,31,32]. To this, a new mating-type system

that is not strictly bipolar or tetrapolar was discovered [30].

Flammulina velutipes, also known as Winter Mushroom and

Enokitake is one of the major cultivated mushrooms in Asia.

Beyond having a tetrapolar mating system with multiple alleles,

the genetics of its mating-type system have remained unknown.

We decided to elucidate the genetic structure of this important

mushroom to map the mating-type genes and use comparative

genomics to understand evolutionary relevant distinctions of the F.

velutipes loci, as well as to implement this knowledge in our

mushroom breeding programs. We obtained a complete map of

the mating-type genes from F. velutipes KACC42780, identified the

specific matA and matB loci and explain some of the events that

caused the significant deviation of the matA region in comparison

to the model species. Mating-type defining genes are currently

used to construct haploid, monokaryotic, mushroom producing

strains and primers for PCR based mating-type identification.

Materials and Methods

Strains and culture conditionsFlammulina velutipes strains were cultivated at 25uC on

100620 mm dishes (SPL, South Korea) containing Potato

Dextrose Agar (HIMEDIA Laboratories, India) for two weeks.

Stocks were transferred to fresh PDA dishes every three months,

sealed (Clean Wrap) and stored at 4uC after three to five days

growth (storage up to one year). For genomic DNA isolation, PDA

was covered with cellophane (general household) prior to

inoculation. For mating experiments, small dishes (60615 mm,

SPL) were inoculated with agar plugs (565 mm) at 0.5 cm from

the center and grown for one week. Fresh plugs from the center of

the plate that contained merged mycelia were subcultured for

another four to eight days prior to clamp formation analysis.

Mushroom cultivation in small bottle cultures was started with

mating of two monokaryons as described above. Confirmed

dikaryons were grown on a 30 mm layer of sterilized sawdust

(Douglas Fir)/wheat bran (Rice) mixture (4:1) in 25645 mm glass

bottles (three mycelial plugs/bottle). After 10 days incubation at

25u the surface of colonized wood was scraped off and bottles were

filled with cold distilled water for 1 hr. Following upside down

drying for 30 min bottles were incubated at 15uC, 95% humidity

for three weeks. Spores were collected by placing the caps of the

mushrooms in small dishes for one day and eluted in sterilized

water prior to plating on PDA. Strains used in this study were

KACC42780 (A3B3), KACC43777 (A4B4) and dikaryon

KACC43778 (A3B3A4B4) available under Korean Agricultural

Culture Collection (KACC), RDA, Korea. Monokaryons for

segregation studies were all derived from dikaryon KACC43778.

For analysis of matB genes in F. velutipes strains from different

geographical locations we used laboratory stock strains 4004-32 &

4004-23 (Korea); 4031-04 & 4031-10 (Korea, commercial strain

Paengi-2); 4028-34 & 4028-38 (Taiwan); 4015-19 & 4015-11

(Japan); 4017-06 & 4017-05 (Korea); 4006-04 & 4006-01 (Korea);

and 4023-01, 4023-05, 4023-29 & 4023-32, F. velutipes var.

longispora (USA).

Genomic DNA extractionFor isolation of genomic DNA, 400 ml of extraction buffer

(100 mM NaCl, 50 mM EDTA, 0.25 M Tris-HCl, 5% SDS),

400 ml of 2x CTAB buffer (2% CTAB, 100 mM Tris-Hcl pH 8.0,

20 mM EDTA pH 8.0, 1.4 M NaCl, 1% polyvinyl pyrrolidone)

and 500 ml phenol-chloroform-isoamylalcohol (25:24:1, Bioneer,

Mating Type Loci in Flammulina velutipes


South Korea) were added to 0.1–0.5 g of lyophilized or fresh

mycelium and briefly vortexed. After 5 min incubation at room

temperature (RT) samples were centrifuged at 13.000 rpm, 4uCfor 5 min. The supernatant was mixed with 0.7 volumes

isopropanol and centrifuged for 10 min, 4uC. After washing with

70% EtOH, air dried samples were eluted in 50–100 ml TE and

treated with RNase A (Bio Basic Inc, Canada) for 30 min at 60uC.

Southern hybridizationDNA probes (500bp) were amplified by PCR (primers; Table

S2). Genomic DNA was digested to completion with SacI at 37uC,

over night. Agarose gels (0.8%) containing the digested gDNA

were soaked in 0.25 M HCl for 20 min for depurination, 30 min

in Denaturation buffer (0.5 M NaOH, 1.5 M NaCl pH 7.5) and

10 min in Neutralization buffer (0.5 M Tris–HCl pH 7.5, 1.5 M

NaCl). DNA was transferred to a nylon membrane (Amersham)

via capillary transfer with 10x SSC (1.5 M NaCl, 0.15 M sodium

citrate) for 20 hrs and membranes were baked at 80uC [33]. Pre-

hybridization and hybridization were done in a hybridization

solution (5x SSC, 50% formamide, 5x Denhardt’s solution, 0.5%

sodiumdodecylsulfate and 25 mg/ml denatured salmon sperm

DNA). Pre-hybridization was performed for 3 hrs and hybridiza-

tion was continued for 24 hrs with a probe labeled with [a-32P]

dCTP (Ladderman TM Labeling kit, Takara, Japan) at 42uC. The

membrane was exposed to an imaging screen (Fuji) for 18 hrs and

DNA bands were visualized using a personal molecular imager

system (Bio-Rad, USA).

PCR analyses and sequencingPCRs were performed using GoTaq flexi kit (Promega Korea

Ltd, South Korea) with specific primers (Bioneer, South Korea) for

each gene (Table S2). Thermal cycling parameters consisted of

10 min denaturation at 94uC, 30 cycles of 94uC, 30 s denatur-

ation, 30 s annealing (temperature 2uC lower than the specific,

lowest melting temperature of primers used), 30 s extension at

72uC and a 10 min final extension at 72uC. Samples for

sequencing were amplified with a BigDye Terminator cycle

sequencing kit (Applied Biosystems, USA), precipitated with 0.1

volume 3 M Sodium Acetate (pH 5.8) and 2.5 volumes 100%

EtOH for 10 min at 270uC and centrifuged for 15 min,

13000rpm at 4uC. Pellets were washed with 70% EtOH, air dried

and eluted in Hi-Di Formamide (Applied Biosystems, USA) and

analyzed on a ABI3730 DNA Analyzer.

Identification of pheromone receptors, pheromoneprecursors and CLA4

A draft genome sequence of Flammulina velutipes monokaryotic

strain KACC42780 was screened for the presence of pheromone

receptors by tblastx searches with the sequence of pheromone

receptor gene Bar1 of Schizophyllum commune (National Center for

Biotechnology Information (NCBI) Genbank accession number

X77949.1), Rcb1 of Pholiota nameko (AB201119.1) and Rcb3 of

Coprinus cinereus (CAA71962). The genome sequence of F. velutipes

will become available at the Agricultural Genome information

Center, RDA (http://10.30.100.11:8000/;) though sequence

downloads and search options may remain restricted. Gene

models for pheromone receptor genes were manually annotated

using tblastx derived homologues from NCBI. Secondary protein

structures of FvSTE3 encoded proteins were predicted using

HMMTOP version 2.0 (Hungarian Academy of Sciences, http://

www.enzim.hu/hmmtop/, [34] and TMHMM Server v.2.0 (CBS,

Denmark, http://www.cbs.dtu.dk/services/TMHMM-2.0/, [35].

All FvSTE3 containing contigs were fully screened for pheromone

precursors. Open reading frames between 40 and 100 amino acids

were manually analyzed for C-terminal CAAX motifs and

screened using Pfam 24.0, http://pfam.janelia.org/, [36]. CLA4

was identified in F. velutipes using CLA4 of Pleurotus djamor (Genbank

AY462110). NCBI Genbank accession numbers for identified

genes of F. velutipes are: FvSTE3.1 = HQ630590; FvSTE3.2/FvPP2/

FvPP3 = HQ630591; FvSTE3.s1 = HQ630592; FvSTE3.s2 = HQ

630593; FvSTE3.s3 = HQ630594; FvSTE3.s4 = HQ630595; Fv

STE3.s5 = HQ630596; FvPP1 = HQ630597.

Identification of homeodomain proteinsThe genome database of F. velutipes (see above) was screened for

the presence of HD1 and HD2 proteins by tblastn searches with

the sequences of C. cinereus 5B6_b1-4 (NCBI Genbank

AF126786.1), C. cinereus (JV6) A42_b2-1 (NCBI Genbank

X79687.1), C. cinereus a42_d1-1 (NCBI Genbank X79688.1), P.

djamor a1-2 (NCBI Genbank AY462112.1), P. djamor a2-2 (NCBI

Genbank AY462112.1), S. commune a_alpha_Y4 (EMBL-Bank

M97181.1) and S. commune a-alpha_Z4 (EMBL-Bank M97181.1).

Additional searches were performed with MIP proteins of P. djamor

(NCBI Genbank AY462112.1) and P. nameko (NCBI Genbank

AB435542.1). Gene models for FvHD1-1, FvHD2-1 and FvHD2-2

were manually annotated. Predicted proteins for FvHD1-1,

FvHD2-1 and FvHD2-2 were analyzed using COILS to identify

dimerization motifs ([37] http://www.ch.embnet.org/software/

COILS_form.html), WoLF PSORT to determine nuclear local-

ization domains ([38] http://wolfpsort.org/) and 9aaTAD

([39,40] http://www.es.embnet.org/Services/EMBnetAT/htdoc/

9aatad/) to search for transactivation domains. NCBI Genbank

accession numbers for the new F. velutipes genes are: FvHD1-1/

FvHD2-2 = HQ630588; FvHD2-1/MIP = HQ630589.

Analysis of the Mat A locusTblastn searches were performed against a draft genome of F.

velutipes (see above), the genome of S. commune strain H4-8,

available at DOE Joint Genome Institute (http://genome.jgi-psf.

org/Schco1/Schco1.home.html) and the genome of Laccaria bicolor

S238N-H82 (available under the Coprinus cinereus project) using

predicted proteins that surround the A mating-type locus of C.

cinerea okayama 7#130, available under the Coprinus cinereus

sequencing project of the BROAD institute of MIT (http://

www.broadinstitute.org/annotation/genome/coprinus_cinereus/

GenomesIndex.html). Gene homologues that were located on F.

velutipes contigs Fv01174, Fv03236 and Fv02632 were used for

comparison in Chromomapper 1.0.10.26 [41].

Phylogenetic analysesPheromone receptor sequences of L. bicolor were obtained from

the Joint Genome Institute (JGI) website according to Niculita-

Hirzel et al. [31]. Sequences of S. commune, P. djamor, C. cinerea and

Cryptococcus neoformans were obtained from NCBI Genbank using

accession numbers as reported by James et al. (Figure 7, in [42]).

Sequences of F. velutipes were obtained as described above. The

protein sequences of pheromone receptors were aligned using

ClustalW 1.64 [43] in combination with Gblocks 0.91b [44] to

eliminate remaining poorly aligned sequences with a setting ‘allow

smaller final blocks’. Ambiguously aligned regions were excluded

from the protein sequences of pheromone receptors in the

phylogenetic analyses. Maximum likelihood analyses were con-

ducted using RAxML 7.04 [45]. A model of PROTGAMMA-

WAGF was selected with an analysis of ProtTest 2.4 [46] and

incorporated in the analysis. Branch values from 1,000 nonpara-

metric bootstrap repetitions were used for nodal support [47]. In

this study, nodes were considered strongly supported when



supported by more than 70% bootstrap value. The resulting tree

was visualized with Figtree 1.3.1.

Results

Identification of pheromone receptor genes andpheromone precursors

A draft genome of Flammulina velutipes monokaryotic strain

KACC42780 (A3B3) was screened for pheromone receptor

homologues and CLA4, a kinase gene shown to be linked to the

B locus in other mushroom species [12]. Six contigs, five with a

single and one with two pheromone receptor homologues were

retrieved. CLA4 was found on a separate contig (Table S1).

Pheromone receptor genes were annotated and named FvSTE3.1,

FvSTE3.2 and FvSTE3.s one to five (FvSTE3.s1 to FvSTE3.s5),

where ‘s’ stands for ‘similar’ to distinguish non mating-type specific

pheromone receptors (see below). Repeated searches with the

newly indentified genes did not uncover additional pheromone

receptors. Six predicted pheromone receptor proteins contained

the seven-transmembrane domains that are characteristic for

fungal pheromone receptors [26]. Alignments of the pheromone

receptor proteins showed that the sequence of FvSte3.s2 deviated

in the region that corresponded with the first transmembrane

region in other pheromone receptors. Moreover, seven amino

acids were deleted in this protein from a highly conserved motif,

truncating its transmembrane sequence (not shown). FvSte3.s2

grouped specifically with FvSte3.s1 in maximum likelihood

analyses (not shown), shared 60% base pair identity and was

separated from gene FvSTE3.s1 by only 6.7 Kb. All together, this

suggests that FvSTE3.s2 is a pseudogene that was derived as a copy

from FvSTE3.s1. Three pheromone precursor genes named FvPP1,

FvPP2 and FvPP3 were identified on contigs with pheromone

receptors. FvPP1 was located 2102bp apart from pheromone

receptor FvSTE3.1 and FvPP2 and FvPP3 flanked pheromone

receptor FvSTE3.2 at distances of 320bp and 427bp respectively

(Figure 1). Notably, FvPp1 contained a C-terminal Tryptophan

(W) behind its CAAX-box (CAAXW). All three FvPp proteins

were classified as fungal pheromones according to Pfam 24.0

searches [36] with E-values of 0.0013, 0.45 and 0.035 for FvPp1,

FvPp2 and FvPp3. Alignment of the FvPp proteins showed that

FvPp2 (54 amino acids) and FvPp3 (46 amino acids) shared 55%

and 54% reciprocal similarity in contrast to FvPp1 (54 amino

acids) with 37%. The three proteins contained a glutamic acid/

arginine (E/R) motif at amino acid position 13/14 (FvPp1) and

15/16 (FvPp2, FvPp3) counted from the C-terminus (Figure S1).

These residues are conserved in various basidiomycetes and have

been speculated to be the site of proteolytic cleavage (see

introduction). Proteolytic splicing at the E/R sites, together with

C-terminal processing, would result in peptides of nine (FvPp1)

and 11 amino acids (FvPp2 and FvPp3) which corresponds well

with the size of other fungal pheromones [19,25].

Segregation analysis of FvSTE3 and FvPP genesIn order to designate the pheromone receptor and pheromone

precursor genes to specific matB loci, their distribution was

analyzed in dikaryon KACC43778, parental strains KACC43777

(A4B4), KACC42780 (A3B3) and two monokaryotic siblings B2

(A3B4) and B27 (A4B3). PCR with specific primers (Table S2)

detected FvSTE3.1 and FvPP1 explicitly in matB3 strains.

Pheromone receptors FvSTE3.2, FvSTE3.s1 to FvSTE3.s5 and

pheromone precursors FvPP2 and FvPP3 were detected in matB3

and matB4 mating-types (Figure 2). Correct amplification and

PCR patterns of FvSTE3 and FvPP genes were confirmed by

sequencing and Southern analysis (Figure 2). Analysis of 16

additional monokaryons with different mating-types showed that

the observed distribution was persistent (Table 1). Monokaryons

were derived as single spore colonies from dikaryon KACC43778

and mating-types were assigned based on clamp formation

patterns (material and methods). Sequence alignment of the

PCR products for each gene that were obtained from the 16

strains revealed two differing copies of genes FvSTE3.s1,

FvSTE3.s2 and FvSTE3.s3. This enabled segregation analysis

through restriction fragment polymorphism of FvSTE3.s1 and

FvSTE3.s2. Subtype FvSTE3.s1-a was invariably detected together

with FvSTE3.s2-d while FvSTE3.s1-b was linked with FvSTE3.s2-c.

Segregation of the couples was independent from the analyzed

mating-type loci (Table 1). Segregation of FvSTE3.2, FvSTE3.s3 to

FvSTE3.s5, FvPP2 and FvPP3 remained undetermined. Taken

together, the results show that only a single pheromone receptor

and pheromone segregate specifically with the B3 mating-type in

F. velutipes KACC42780.

Polymorphism of pheromone receptorsMating-type specific genes in fungi are characterized by highly

polymorphic alleles. We examined the distribution and polymor-

phism of the FvSTE3 genes in various F. velutipes strains, from

different countries (materials and methods). Genomic DNA, of two

compatible monokaryons for each strain, was analyzed by PCR

with specific primers (Table S2). FvSTE3.s1, FvSTE3.s2, FvSTE3.s3

and FvSTE3.s4 were frequently amplified with six, five, seven and

seven products out of 10 strains. FvSTE3.1, FvSTE3.2 and

FvSTE3.s5 were not detected outside the control (Table 2). New

primer sets (Table S2, *additional sets) resulted in frequent

amplification of FvSTE3.s5 but FvSTE3.1 and FvSTE3.2 remained

undetected. Sequences for the amplified genes shared respectively

92–99 percent base pair similarity (Table 2). Low polymorphism of

FvSTE3.s1 to FvSTE3.s5 showed that those genes are not mating-

type specific pheromone receptors; their amino acid sequences are

identical. The absence of FvSTE3.1 and FvSTE3.2 from all tested

F. velutipes strains, especially in comparison with the other

pheromone receptors, clearly indicates polymorphic alleles for

those two genes and therefore supports a mating-type specific role.

Figure 1. Organization of the matB3 locus. The figure depicts arepresentative map of scaffold 1 with an enlarged section (black) for thematB3 locus. Pheromone receptor genes are almost evenly spaced andcover one third of scaffold 1. FvSTE3.1 and FvPP1 specifically segregatewith matB3a (shaded dark grey) and are indicated by red arrows.Pheromone receptor FvSTE3.2 and pheromones FvPP2 and FvPP3 (lightred arrows) presumably form a second, matBb sublocus (shaded lightgrey). Non mating-type specific pheromone receptors FvSTE3.s3 andFvSTE3.s5 (grey arrows) might be part of the matB locus as well.doi:10.1371/journal.pone.0022249.g001



Phylogenetic analyses of pheromone receptorsMaximum likelihood analyses based on the protein sequences of

F. velutipes pheromone receptors and that of other basidiomycete

species indicated a division into two distinct clades (Figure 3, Clade

A and B). The four pheromone receptors FvSte3.s1, FvSte3.s3,

FvSte3.s4 and FvSte3.s5 formed a separate, strongly supported

clade within clade B (Figure 3, purple shaded). FvSte3.2 was also

grouped in clade B with strong support, together with pheromone

receptor ScBbr2 from S. commune. In turn, they were strongly

grouped with CcRcb2 and LbSte3.1 from C. cinerea and L. bicolor

(Figure 3, yellow shaded). ScBbr2, CcRcb2 and LbSte3.1 are all

mating-type specific pheromone receptors [18,31]. FvSte3.1 was a

Figure 2. Distribution of pheromone receptors and pheromones in matB3 and matB4 loci. Southern blots confirmed the PCR distributionpatterns of the pheromone receptors and pheromones in dikaryon KACC43778; A3B3A4B4 (DK), strain KACC43777; A4B4, (B18), KACC 42780; A3B3,(B20) and two monokaryotic siblings 4019-B2; A3B4, (B2) and 4019-B27; A4B3, (B27). Pheromone receptor FvSTE3.1 and pheromone FvPP1 areexclusively detected in strains containing the matB3 locus (B20 and B27), and as a single copy in the dikaryon (DK). The small, equally sized signals forFvSTE3.s4 on the Southern blot are caused by an internal SacI restriction site in this gene. PCR detection patterns for FvSTE3.1, FvSTE3.2 and FvSTE3.s1are inserted as panels in the bottom of the Southern analyses. Strains that contained a copy of a gene invariably generated a specific PCR product. Nofalse products were amplified with our primers. Pheromone receptor FvSTE3.2, FvSTE3.s1 to FvSTE3.s5 and pheromone receptors FvPP2 and FvPP3were detected in matB3 and matB4 mating-types.doi:10.1371/journal.pone.0022249.g002



Table 1. Distribution of pheromone precursors, pheromone receptors and homeodomain genes in F. velutipes monokaryoticsiblings.

Strain 4019- DK B10 B11 B13 B17 B19 B21 B22 B23 B25 B28 B30 B32 B34 B35 B36 B38

A mating-type A3A4 A4 A4 A3 A3 A4 A4 A3 A4 A3 A3 A3 A4 A4 A3 A4 A4

B mating-type B3B4 B4 B3 B3 B3 B3 B3 B4 B4 B3 B3 B4 B4 B3 B4 B4 B4

Gene:

FvSTE3.1 + 2 + + + + + 2 2 + + 2 2 + 2 2 2

FvSTE3.2 + + + + + + + + + + + + + + + + +

FvSTE3.s1 + + + + + + + + ? + + + + + ? + +

RFLP s1-a/b b a a a b b a ? a a b a a ? a a

FvSTE3.s2 + + + + + + + + + + + + + + + + +

RFLP s2-c/d c d d d c c d d d d c d d c d d

FvSTE3.s2 + + + + + + + + + + + + + + + + +

FvSTE3.s3 + + + + + + + + + + + + + + + + +

FvSTE3.s4 + + + + + + + + + + + + + + + + +

FvSTE3.s5 + + + + + + + + + + + + + + + + +

Fvpp1 + 2 + + + + + 2 2 + + 2 2 + 2 2 2

Fvpp2 + + + + + + + + + + + + + + + + +

Fvpp3 + + + + + + + + + + + + + + + + +

FvHD1-1 + + + + + + + + + + + + + + + + +

FvHD2-1 + 2 2 + + 2 2 + 2 + + + 2 2 + 2 2

FvHD2-2 + 2 2 + + 2 2 + 2 + + + 2 2 + 2 2

Distribution of pheromone receptors and pheromones was analyzed in monokaryotic F1 progeny from dikaryon KACC43778 (DK). Presence of genes is indicated by ‘+’and absence by ‘-‘. FvSTE3.s1 and FvSTE3.s2 show the distribution of their respective, non digested PCR products. Their subtypes, genes that were identified in sequenceanalysis of FvSTE3.s1 and FvSTE3.s2 are RFLPs1-a (a), RFLPs1-b (b), RFLPs2-c (c) and RFLPs2-d (d). Subtypes were detected by restriction length polymorphism analysis ofPCR products. The table shows specific linkage of pheromone receptor FvSTE3.1 and pheromone FvPP1 with the matB3 locus. Copies of FvSTE3.s1 and FvSTE3.s2recombine in pairs but independent from the mating-type loci; FvSTE3.s1-a is coupled to FvSTE3.s2-d and FvSTE3.s1-b is coupled to FvSTE3.s2-c. Homeodomain genesFvHD2-1 and FvHD2-2 co-segregate specifically with the matA3 locus. Segregation of FvSTE3.2, FvSTE3.s3, FvSTE3.s4, FvSTE3.s5, FvHD1-1, FvPP2 and FvPP3 could not bedetermined. Copies of those genes were detected in all mating-types.doi:10.1371/journal.pone.0022249.t001

Table 2. Presence of pheromone receptors in different F. velutipes strains of wide geographical distribution.

Strain Pheromone receptor

FvSTE3.1 FvSTE3.2 FvSTE3.s1 FvSTE3.s2 FvSTE3.s3 FvSTE3.s4 FvSTE3.s5

DK + + + + + + +

4004 23 2 2 2 + + + +

4004 32 2 2 + + + + +

4031 4 2 2 + + 2 2 +

4031 10 2 2 2 + 2 + 2

4028 34 2 2 + + + + +

4028 38 2 2 2 2 + 2 +

4023 23 2 2 + 2 + + 2

4023 29 2 2 + 2 + 2 +

4023 1 2 2 2 2 + + +

4023 5 2 2 + 2 2 + 2

Similarity 2 2 93–97% 92–98% 97–98% 93–99% 96–99%

The table shows a clear distinction in distribution between FvSTE3.1, FvSTE3.2 and other pheromone receptors from strain KACC42780 when analyzed in geographicallydistant F. velutipes strains. Presence was determined by PCR analysis on genomic DNA of two compatible mating types of each strain. Locations of the strains were asfollows: 4004223 +4004232, Korea; 403124+4031210, Korea; 4028234+4028238, Taiwan; 402321+402325 & 4023223+4023229, F. velutipes var. longispora, USA.FvSTE3.1 and FvSTE3.2 are uniquely amplified from the control, dikaryon KACC43778. All other pheromone receptors are regularly amplified from geographically distantF. velutipes strains. Sequence similarity of PCR products for each regularly amplified pheromone receptor was high showing that they are non mating-type specific. Basepair similarity is shown in percentage in the bottom line.doi:10.1371/journal.pone.0022249.t002



close relative of ScBbr1 from S. commune in clade A and both were

strongly grouped with LbSte3.2 and LbSte3.5 of L. bicolor which,

except for LbSte3.5, are mating-type specific pheromone receptors

as well. Close grouping of FvSte3.1 and FvSte3.2 with other

known mating-type specific pheromone receptors formed a strong

indication that they were also mating-type specific pheromone

receptors. Interestingly, ScBBR1 and ScBBR2 are alleles at the

same matBb locus of S. commune [18,23].

Structure of the matB3 locusThe pheromone receptor and pheromone precursor genes

together with CLA4 were mapped based on linkage of their

respective contigs to a draft genome of F. velutipes KACC42780.

The MatB3a locus containing FvSTE3.1 and FvPP1 was located on

scaffold 1 and flanked upstream by FvSTE3.s3 and FvSTE3.s5, and

downstream by FvSTE3.2, FvPP2 and FvPP3 (Figure 1). Notably,

the pheromone receptors were almost evenly spaced by 177 Kb,

181 Kb and 184 Kb which was considerably more distant than

has been reported for other basidiomycetes [31,32]. Localization

of the genes on the same fragment of scaffold 1 suggested that they

might be part of the matB3 locus. FvSTE3.s1 and FvSTE3.s2 were

mapped on scaffold 26 and FvSTE3.s4 on scaffold 29 (not shown).

CLA4 was located at 0.98 Mb and 1.16 Mb distance from the

borders of scaffold 8 (not shown). This demonstrated that

FvSTE3.s1, FvSTE3.s2, FvSTE3.s4 and CLA4 were not linked to

the matB3 locus.

Identification of homeodomain genesOne F. velutipes HD1 gene (FvHD1-1), two HD2 genes (FvHD2-

1, FvHD2-2) and the Mitochondrial Intermediate Peptidase (MIP) gene

were identified in the F. velutipes genome located on a single contig;

Fv01174 (Figure 4, Table S1). MIP was included because this gene

is closely linked to HD genes in all Agaricomycetes [11,12].

Analysis of the FvHD2-1 and FvHD2-2 gene models (accession

codes in materials and methods) showed intron-exon distributions

and long C-terminal exons similar to that of homeodomain gene

a2-1 and b2-1 of C. cinerea [14]. The first and the second intron

interrupted the homeodomain at the exact same locations in F.

velutipes and C. cinerea. The second introns have been reported to be

conserved in several other basidiomycetes as well [7,14,48,49].

FvHD1-1 (six predicted introns) showed no intron-exon resem-

blance to C. cinerea HD1 genes associated with a2-1 and b2-1. No

reliable coiled coils (less than 20–30% drops in the probability

between weighted/non-weighted analysis) that could indicate

dimerization motifs were found in the F. velutipes HD genes but

nine-amino acid transactivation domains were detected in the N-

termini of FvHd1-1, FvHd2-1 and FvHd2-2 (Table S3). These

domains are generally found in mammalian and yeast transcrip-

Figure 3. Phylogenetic tree of mating-type specific and non mating-type specific pheromone receptor protein sequences. The treeshows phylogeny amongst pheromone receptor proteins from F. velutipes (Fv), C. cinerea (Cc), L. bicolor (Lb), S. commune (Sc), P. djamor (Pd) and C.neoformans (Cn). Nodal supports with more than 70% bootstrap values are considered strongly supported and displayed in the tree. Known mating-type specific pheromone receptors are depicted in blue. Two major clades are distinguished, labeled A and B. The four non mating-type specificpheromone receptors of F. velutipes FvSte3.s1 to FvSte3.s5 (pseudogene FvSTE3.s2 was excluded) form a separate group (shaded purple) within cladeB that is supported by strong branch values. The clade including FvSte3.1 and two other known mating-type specific pheromone receptors is shadedin orange. The clade that contains FvSte3.2 is shaded in yellow. Both these clades are supported by strong branch values. FvSte3.1 and FvSte3.2group closest with SCBbr1 and SCBbr2, respectively. Clades that contain known mating-type specific pheromone receptors are strong evidence formating-type specificity of other clade members. LbSte3.5 is a notable exception.doi:10.1371/journal.pone.0022249.g003



tion factors [40]. FvHd1-1, FvHd2-1 and FvHd2-2 were further

predicted to contain single and or bipartite nuclear localization

signals (Table S3).

Segregation of FvHD genesIn order to link the FvHD genes to specific matA loci we

determined their distribution in matA3 and matA4 mating-types.

Specific primers for FvHD2-1, FvHD2-2 and FvHD1-1 were

designed for exon regions that flanked the conserved homeodo-

main (Table S2). FvHD2-1 and FvHD2-2 showed specific linkage

to the matA3 locus whereas FvHD1-1 was detected both in strains

with matA3 and matA4 loci (Table 1). PCR products (average

450bp) obtained from matA3 and matA4 strains with FvHD1-1

specific primers, showed 97% base pair similarity and 100%

amino acid identity (no polymorphism) demonstrating presence of

a copy instead of different alleles. This means that the matA3 locus

lacks a specific HD1 gene in comparison with matA4.

Structure of the matA locusMIP was located 201bp distant from FvHD2-1 which is in

contrast to most known basidiomycetes where MIP is directly

flanked by a HD1 gene that is part of a HD1/HD2 couple [12].

Detailed screening of the sequence adjacent to MIP and FvHD2-1

did not reveal a FvHD1 gene but instead, a hypothetical gene with

no orthologues near matA subloci in L. bicolor (373Kb distant), C.

cinerea (different chromosome) and S. commune (absent). Gene

FvHD1-1 and FvHD2-2 that are 199bp apart, showed divergent,

outward transcription directions (Figure 5) which is typical for

HD1/HD2 couples in basidiomycetes [12]. FvHD1-1 and FvHD2-

2 were separated by 73 Kb from MIP and FvHD2-1. MatA subloci

at large distance have so far only been reported for S. commune

[4,32]. No Beta-flanking gene was detected near the HD genes, which

was unusual [12]. These deviations showed that the matA locus

had been subjected to several rearrangements relative to all other

Agaricomycetes.

Synteny of matA regionsTo unravel the events that altered the organization of the matA

region of F. velutipes we mapped synteny of 200 successive genes

surrounding the matA locus of C. cinerea (chromosome 1,

bp2471986-2934538) with that of L. bicolor, F. velutipes and S.

commune. The latter was included because this fungus also has

separated matA subloci and is taxonomically closer to F. velutipes

than C. cinerea and L. bicolor [50]. Since there had been no

annotation of the F. velutipes genome, C. cinerea genes were

acknowledged as ‘‘syntenic’’ in F. velutipes when similar protein

sequences with expect values equal to or smaller than 1029 were

obtained by tblastn. Loss of syntenic genes in the matA region

from C. cinerea was indicated by 11 unique genes, and 21 (L.

bicolor), 17 (F. velutipes) and 21 (S. commune) genes from C. cinerea

that were not detected in one or two of these species. The selected

C. cinerea region covered the matA containing F. velutipes contig

Fv01174 (304 Kb) as well as parts of contig Fv02632 and

Fv03236, the last which was found to contain the missing Beta

flanking gene. The three contigs were linked to scaffold 3 (3.8 Mb)

of the F. velutipes draft genome where Fv01174 was flanked

upstream at 2.1 Kb by Fv03236 and downstream by Fv02632 at

28 Kb (Figure 4). Our analysis identified synteny between C.

Figure 4. Synteny in matA regions of F. velutipes, C. cinerea, L. bicolor and S. commune. Syntenic mapping of the matA regions from F.velutipes (Fv), C. cinerea (Cc) and S. commune (Sc) in Chromomapper [41] reveals significant differences in gene arrangement. Distances are shown inkilo base pair (Kb) and F. velutipes contigs are depicted above the F. velutipes gene bar. A comparison of C. cinerea and L. bicolor (Lb) is shown in‘panel A’ demonstrating the high synteny between those species. The black line under the gene bar of C. cinerea and above L. bicolor marks the350 Kb, high syntenic region of in the figure and in panel A. MatA loci of all species are indicated by green boxes. Rearranged regions in F. velutipesand S. commune are indicated by blue and purple lines respectively. Dashed light purple lines in S. commune indicate non syntenic regions. Borders ofthe highly syntenic region can be clearly recognized by yellow genes in C. cinerea and coincide both with borders of rearranged gene clusters in F.velutipes and S. commune and with individually rearranged genes in L. bicolor (panel A). MatA loci (green) and a second spot (red triangle, red bars inpanel A) are conserved sites of recombination in all four species. The gene order in F. velutipes and S. commune is changed by different events. Fvelutipes shows many inversions of gene clusters when compared to C. cinerea. S. commune shows rearrangements of larger, different gene groups.The overall gene orders of F. velutipes and S. commune are very similar to that of C. cinerea and L. bicolor which strongly suggests that these lattermodels represents an ancestral organization.doi:10.1371/journal.pone.0022249.g004



cinerea and L. bicolor over a remarkably large segment of 350 Kb

(Figure 4, panel A), much larger than previously demonstrated

[31]. In addition, we identified the boundaries of this syntenic

segment, which are clearly denoted by genes from C. cinerea that

have no local homologue in L. bicolor, were differently distributed

in L. bicolor, or were inverted (Figure 4, panel A). Those

boundaries coincide with the ends of the syntenic part of the

matA region from F. velutipes (Figure 4). The matA region in F.

velutipes showed highest synteny over the first 250 Kb of the

350 Kb segment and somewhat lower in the last 100 Kb

(Figure 4, black line, 250 Kb; start to red mark, 100 Kb; red

mark to end). Though the specific gene order of F. velutipes in

comparison to C. cinerea was changed by inversions, the overall

gene order as found in C. cinerea was shown to be strongly

preserved. None of the inverted gene groups was translocated

(Figure 4). In S. commune the syntenic 350 Kb matA region was

also recovered, albeit in three different sections with 140 Kb and

200 Kb interval distances (Figure 4, purple and dashed light

purple lines). Synteny was highest in the two segments

corresponding to the first 250 Kb of the 350 Kb matA region

and lower for the third fragment, resembling F. velutipes.

Moreover, synteny of the fragments in S. commune ended at the

same relative locations that represented the synteny boundaries in

C. cinerea, L. bicolor and F. velutipes. Finally, the gene order within

the three segments was strongly conserved, indicating a high level

of gene conservation in this entire region for all Agaricomycetes.

Inside the 350 Kb segment, two spots showed high recombina-

tion in all four species. The first spot was the matA locus itself, the

second spot formed a small gap in synteny located between C.

cinerea genes CC1G_01875.3 and CC1G_01877.3 (both hypothet-

ical genes) and their respective orthologues. Exemplary, one of

the few genes that was repositioned in L. bicolor in comparison to

C. cinerea was inserted in this gap. In F. velutipes and S. commune

gene clusters were separated by inversion or translocation at both

of these locations (Figure 4, 5, 6). Detailed comparison of the F.

velutipes gene order to that of C. cinerea showed that the matA locus

of F. velutipes was separated by inversion of two 70 Kb fragments

directly left and right of FvHD1-1 and FvHD2-2 (Figure 4, 5).

Modeling a reversion of those clusters reunited the Beta-flanking

gene, FvHD2-1, MIP, FvHD1-1 and FvHD2-2 in a similar

distribution as found in C. cinerea and L. bicolor. Notably,

FvHD2-2 and FvHD1-1 were never moved during the rearrange-

ments (Figure 5). Synteny mapping also revealed important

differences between the matA regions of F. velutipes and S. commune.

First, all S. commune homeodomain genes were repositioned,

whereas FvHD2-2 and FvHD1-1 retained their position during

rearrangements. This shows that the respective matA loci were

split between different genes. Second, the S. commune fragments

representing the first part of the high syntenic 350 Kb region,

and whose rearangement caused separation of the matA subloci,

constituted 140 Kb and 260 Kb (Figure 4, purple lines). This was

significantly larger than the 70 Kb in F. velutipes (Figure 4, blue

lines) and shows that different fragments were rearranged. Third,

the S. commune fragments were wedged by sections not syntenic to

the 350 Kb C. cinerea matA region (Figure 4, dotted light purple

lines), which was not observed in F. velutipes.

To specify the events that split the S. commune matA locus we

extended the synteny map between C. cinerea and S. commune in

both directions, using on average each next 5th gene on the C.

cinerea genome (Figure 6). Model A (Figure 6A) showed highest

similarity to the gene order in C. cinerea, with all major S. commune

gene clusters in an equal orientation. Model B (Figure 6B) with the

S. commune chromosome in an opposite orientation, showed

inversion of all but two major gene clusters. Both models indicated

transposition of two (model A) or three (model B) high syntenic

fragments corresponding to the 350 Kb region (Figure 6, blue and

purple lines respectively). In addition, the two transposed

fragments in model A were exchanged in order in comparison

to C. cinerea. Repositioning of the two fragments representing the

high syntenic 250 Kb region (discussed above) in model A,

between the genes flanking the 350 Kb region and the genes

representing the last 100 kb of the 350 Kb region on the other side

(Figure 6A, C, black lines), reconstituted a matA region mainly

identical to C. cinerea. Model B clearly indicated inversions as a

cause for separation of the S. commune matA subloci. Indeed

reversion of the genes underscored by the first (390 Kb) and

second (210 Kb) purple line reunited the matA subloci (Figure 6D).

However, the S. commune matA region that resulted from those

inversions still contained a large non syntenic fragment (Figure 6,

red dotted line) and differed considerably in gene order from C.

cinerea.

Figure 5. Gene order near the matA locus of F. velutipes before and after inversion. A detailed overview of the synteny map in Figure 4,shows the individual genes of the matA loci from F. velutipes and C. cinerea. Homeodomain gene FvHD2-1 (green), the Beta flanking gene (BFG, purple)and the Mitochondrial Intermediate Peptidase (MIP, blue) are presently separated, flanking both sides of the FvHD1-1/FvHD2-2 gene couple (bottomgene bar). The top gene bar shows the ancestral gene order in F. velutipes. Apart from a different number of HD genes, the matA locus is identical tothat of C. cinerea. The synteny map clearly demonstrates that two inversions have caused separation of the matA subloci in F. velutipes. Severaladditional inversions are indicated in the ancestral F. velutipes locus. Notably, FvHD2-2 and FvHD1-1 maintained their position during all changes.doi:10.1371/journal.pone.0022249.g005



Discussion

With the fast growing range of sequenced genomes, new

information of mating-type genes quickly becomes available.

However, complete analyses of the genetic structure of mating-

type loci in mushroom forming fungi remain rare. Our study

focused on the genetic structure of the mating-type loci from

Flammulina velutipes (Winter Mushroom or Enoki). This led to the

finding of several characteristics for F. velutipes as well as new facts

that will help gaining insight in mating-type locus evolution,

especially in Agaricales.

The B mating-type genesThree pheromone precursors were identified near pheromone

receptors in F. velutipes. As we limited our screen for pheromones to

the six contigs that contained pheromone receptors (total 446 Kb),

more pheromones can be expected to be found in the genome

[31]. However, those pheromones will not be closely associated

with the pheromone receptors and are unlikely to perform a

mating-type specific role. FvPp1, the only pheromone precursor

that was specifically linked to the matB3 mating type is unique,

ending with a tryptophan (W) after the CAAX-box. We suggest

that this extra tryptophan originates from the former STOP codon

(Stop = TAA, TAG or TGA). A single base pair substitution would

suffice to change TAG or TGA to TGG (TGG = W). It is

uncertain if the additional tryptophan impairs pheromone

processing; yet deviant farnesylation signals have previously been

demonstrated to be functional [23].

We identified two mating-type specific pheromone receptors

(accompanying pheromones, phylogenetic grouping with other

mating-type specific pheromone receptors and high sequence

polymorphism) named FvSTE3.1 and FvSTE3.2, which represent-

ed two of the three fungal pheromone receptors families. No genes

of the third pheromone receptor family were detected in the

matB3 strain, though this family might be expected in other F.

velutipes matB loci.The identification of multiple non mating-type

specific pheromone receptors in F. velutipes is in line with recent

discoveries in L. bicolor, C. cinerea and S. commune [3,31,32]. We

demonstrated a clear distinction between the two pheromone

receptor types based on phylogenetic distribution and sequence

polymorphism. Different sequence polymorphism indicates that

these genes are subjected to different selection mechanisms.

Somewhere in evolution, non mating-type specific pheromone

receptors must have been functionally and selectively separated

from mating-type specific ones. At the moment it is uncertain if

non mating-type specific pheromone receptors are functional and

what role they perform in the fungus. At least, their role is not

mating-type specific (this study, [31,32]).

Structure of the matB locusMating-type specific pheromone receptor FvSTE3.2, together

with pheromone precursor FvPP2 and FvPP3, was positioned in

the same region of scaffold 1 as the matB3a locus containing

FvSTE3.1 and FvPP1 (Figure 3). Arguably, FvSTE3.2 and the

accompanying pheromone precursors comprise a functional

Figure 6. Synteny modeling of the events that separated the matA subloci of S. commune. Panel A and B show synteny maps of C. cinereaand S. commune, with the latter in opposite orientations. Map A shows high gene order similarity whereas B suggests inversion of nearly all major S.commune segments. In map A, two large segments (blue lines) that correspond to the highly conserved 350 Kb matA region (Figure 4) are displacedfrom the region flanked by the black lines. Repositioning of those fragments (in reversed order) reconstitutes the genomic organization also found inC. cinerea (C) with a single matA locus (green) and a continuous high syntenic 350 Kb matA region. Map B clearly shows two inversed gene clusters(purple lines) that have separated the S. commune matA locus. Repositioning of those clusters (map D) reunites the matA locus, but does not reunitethe 350 Kb region. About one third (purple line most right, map D) of the genes corresponding to this region remains separated by a large nonsyntenic segment (,250 Kb, red dotted line) and most other S. commune gene clusters still display inversed orientation in comparison to the geneorder in C. cinerea.doi:10.1371/journal.pone.0022249.g006



second sublocus (yet our strain is just homozygous) since this

specific receptor was not detected in any of the F. velutipes strains

from different locations. The 177 Kb distance between both

subloci makes the matB locus of F. velutipes exceptionally large in

comparison to other higher basidiomycetyes [20,31,32]. To this,

non mating-type specific pheromone receptors have been

demonstrated to be linked to matB loci [3,32] meaning that

FvSTE3.s3 and FvSTE3.s5 could be part of matB3 as well. This

would increase the matB3 locus to over 500 Kb. Identification of

additional F. velitupes matB loci should reveal if this large distance is

consistent in F. velutipes and if FvSTE3.s3 and FvSTE3.s5 are truly

connected.

The presence of FvSTE3.1 and FvSTE3.2 on different subloci

corresponds with their phylogenetic separation into two clades that

were derived through duplication [12]. It is surprising that their

closest homologues, ScBBR1 and ScBBR2, are alleles on the same

locus (matBb) in S. commune [18,23]. The diversity between F.

velutipes and S. commune, as well as the large distance in the first,

shows that the genomic organization of matB loci should be

considered more flexible than previously has been assumed.

The A mating-type genesThree homeodomain genes distributed over two distant subloci

were identified in F. velutipes. Both subloci were specifically linked

to mating-type A3 (FvHD2-1 and FvHD2-2) yet the only present

HD1 gene (FvHD1-1) was found in matA3 and matA4 mating

types. MatA4 thus contains a different matAa allele, consisting of

FvHD1-1 either combined with another HD2 gene than FvHD2-2

or as a single gene. This means that FvHD1-1 or FvHD2-2, that

form a mating-type gene couple in matA3, have been indepen-

dently recombined in other mating types. Lack of a matA3 specific

HD1 gene dictates existence of at least one other HD1 gene in

matA4, presumably in the matAb locus.

Structure of the matA regionWe mapped the synteny between F. velutipes, S. commune, L. bicolor

and C. cinerea based on blast searches with successive genes of the

latter. Consequently, few genes that are missing in C. cinerea but that

might be syntenic between other species remained undetected. The

high detail of our maps however, showed that the applied method was

accurate for both annotated and non annotated species. F. velutipes

and S. commune showed different gene orders when compared, yet

both followed the overall gene order of C. cinerea and L. bicolor. This,

together with non separated matA loci in C. cinerea and L. bicolor shows

that they represent an ancestral organization. We identified a 350 Kb

matA region that is strongly conserved amongst Agaricales. Notably,

synteny of genes belonging to this region is preserved even if parts

become separated by chromosomal rearrangements as was shown in

S. commune. The borders of the 350 Kb region, as well as two internal

hotspots (one of which is the matA locus) are conserved sites of

recombination in two major clades of the Agaricales as classified by

Matheny et al. [50]. They mark the edges of important rearranged

segments in F. velutipes as well as S. commune. Reasonably, one might

expect rearrangements of the strongly conserved 350 Kb matA

region in other Agaricales, especially in the Marasmioid clade that

contains F. velutipes and S. commune.

Our segregation experiments showed that both matA3 subloci

are linked despite their 73 Kb distance. Until now, far distant

matA subloci in Agaricales were only known in S. commune

(,450 Kb) and considered to be an exception. The matA subloci

of F. velutipes were demonstrated to be separated by inversion of

two (,70 Kb) gene clusters and are clearly derived from an

ancestral locus as represented by C. cinerea. It has been generally

assumed that the matA subloci of S. commune have followed a

similar course leading to separation. However, our analysis

showed significant differences between rearrangements in F.

velutipes and S. commune. What is more, the analyses of the extended

synteny map between S. commune and C. cinerea strongly indicate

that the matA subloci of S. commune were separated by transposition

of gene clusters instead of inversions. As shown in model A

(Figure 6A, C) a two step transposition of two high syntenic

segments corresponding to the conserved 350 Kb matA region,

directly results in a gene order mostly similar to that in C. cinerea.

Though it is possible to reconstitute the ancestral matA locus by

two inversions of considerably larger fragments as shown in model

B, many rearrangements including transpositions remain

(Figure 6B and D). Both the smaller sizes of the rearranged

fragments and the fewer steps needed to reconstitute the ancestral

gene order support the transposition model.

F. velutipes provides a phylogenetically diverse species with an

unusual mating type system (one of two components of the matA3

locus, HD2, is variable relative to matA4 while the HD1 is

identical and matB3 contains a unique pheromone precursor). This

enables comparative genomics to identify trends in mating type

locus evolution. Studies of mating-type genes in Agaricales have

shown that subloci are typically closely linked (10–20 kb) with the

directly surrounding genes, especially in the matA locus, being

highly syntenic. In our study we determined that the Agaricales in

fact have very large scale synteny at matA (,350 Kb) and that this

synteny is maintained even when parts of this region are separated

through chromosomal rearrangements (S. commune). Four conserved

recombination hotspots allow reshuffling of large fragments of this

region which resulted in separation of the matA subloci of F. velutipes

as well as of S. commune, by different events. This implies that

separation of matA loci is not exceptional and might be expected in

other Agaricales. In addition to matA, we determined that also matB

loci can exist over large distances (,180 Kb) and that non mating-

type specific pheromone receptors and mating-type specific ones are

controlled by different selection mechanisms. Finally, the genes that

were linked to specific mating types will serve as important

molecular markers for breeding.

Supporting Information

Figure S1 Alignment of the three pheromone precursorproteins from F. velutipes KACC42780. Conserved amino

acids in the pheromone sequences are marked with *. The

conserved putative proteolytic site represented by E and R is

indicated by a bold line. The CAAX-box of each protein is

designated by a dashed line. The C-terminal halves of FvPp2 and

FvPp3 are highly similar (boxed).

(TIF)

Table S1 Contigs of F. velutipes KACC42780 that wereused in this study. The table enlists the contigs of F. velutipes

KACC42780 that were used in this study, describing their

respective size and important genes that are located on those

contigs. Genes that were specifically linked to mating type loci in

this study are indicated with *. Pseudogene FvSTE3.s2 is italicized.

The STE3 prefix ‘Fv’ is added for Flammulina velutipes, the small ‘s’

preceding STE3 numbers is added to distinguish non mating-type

specific pheromone receptors (FvSte3 similar) from mating-type

specific pheromone receptors. Gene accession numbers for

pheromone receptors, pheromones, homeodomain genes and

MIP are given in the material and methods.

(DOC)

Table S2 Primers used in this study. List of specific primers

for each pheromone receptor, pheromone and homeodomain



gene from F. velutipes KACC42780. The first primer sets for each

gene were used to determine distribution of genes for segregation

analysis. The ‘*additional sets’ for FvSTE3.1, FvSTE3.2 and

FvSTE3.s5 were newly designed after amplification in additional F.

velutipes strains failed. Primer ‘Fl matA 1-1 rv 500bp’ for gene

FvHD1-1 is exceptional and anneals in an intron.

(DOC)

Table S3 Amino acid positions of domains that weredetected in proteins FvHd1-1, FvHd2-1 and FvHd2-2.Homeodomain proteins were analyzed for 9 amino acid

transactivation domains (9AA TAD) with 9aaTAD [40] and

nuclear localization signals (NLS) by WoLF PSORT [38].

Numbers in the table refer to the amino acid (AA) numbers of

the domains in the respective proteins.

(DOC)

Acknowledgments

We would like to thank Jeong-Hun Baek of Macrogen Inc. from the Next

Generation Genomics Division in Seoul for his help with analyzing the

sequence data from the draft genome of F. velutipes. We are also grateful to

the Agricultural Genome information Center of the Rural Development

Administration for access to the F. velutipes genome sequence before

publication.

Author Contributions

Conceived and designed the experiments: W-SK AFvP B-ML. Performed

the experiments: AFvP S-YP P-GS K-YJ Y-BY YJP. Analyzed the data:

AFvP W-SK G-HS TYJ. Contributed reagents/materials/analysis tools:

W-SK AFvP Y-BY Y-JP. Wrote the paper: AFvP G-HS.

References

1. Kues U (2000) Life History and Developmental Processes in the Basidiomycete

Coprinus Cinereus. Micr Mol Biol Rev 64: 316–353.

2. Casselton LA, Challen MP (2006) The Mating Type Genes of the Basidiomycetes. In:

Kues U, Fischer R, eds. The Mycota. Springer: New York. pp 357–374.

3. Raudaskoski M, Kothe E (2010) Basidiomycete Mating Type Genes and

Pheromone Signaling. Eukar Cell 9: 847–859.

4. Raper JR (1966) Genetics of Sexuality in Higher Fungi. New York. Ronald press.

5. Casselton LA, Olsenicky NS (1998) Molecular Genetics of Mating Recognition

in Basidiomycete Fungi. Microbiol Mol Biol Rev 62: 55–70.

6. Brown AJ, Casselton LA (2001) Mating in mushrooms: increasing the chances

but prolonging the affair. TRENDS in Gen 17: 393–400.

7. Stankis MM, Specht CA, Yang H, Giasson L, Ullrich RC, et al. (1992) The Aa

mating locus of Schizophyllum commune encodes two dissimilar multiallelic

homeodomain proteins. Proc Natl Acad Sci U S A 89: 7169–7173.

8. Badrane H, May G (1999) The divergence-homogenization duality in the evolution

of the b1 mating type gene of Coprinus cinereus. Mol Biol Evol 16: 975–986.

9. Halsall JR, Milner MJ, Casselton LA (1999) Three subfamilies of pheromone

and receptor genes generate multiple B mating specificities in the mushroom

Coprinus cinereus. Genetics 154: 1115–1123.

10. May G, Shaw F, Badrane H, Vekemans X (1999) The signature of balancing

selection: fungal mating compatibility gene evolution. Proc Natl Acad Sci U S A

96: 9172–9177.

11. James TY, Kues U, Rehner SA, Vilgalys R (2004a) Evolution of the gene

encoding mitochondrial intermediate peptidase and its cosegregation with the A

mating-type locus of mushroom fungi. Fungal Gen Biol 41: 381–390.

12. James TY (2007) Analysis of mating-type locus organization and synteny in

mushroom fungi- beyond model species. In: Heitman J, Kronstad J, Taylor JW,

Casselton LA, eds. Sex in fungi: molecular determination and evolutionary

implications. Washington DC: ASM Press. pp 317–331.

13. Tymon AM, Kues U, Richardson WVJ, Casselton LA (1992) A fungal mating

type protein that regulates sexual and asexual development contains a POU-

related domain. EMBO 5: 1805–1813.

14. Kues U, Asante-Owusu RN, Mutasa ES, Tymon AM, Pardo EH, et al. (1994)

Two classes of homeodomain proteins specify the multiple a mating types of the

mushroom Coprinus cinereus. Plant Cell 6: 1467–1475.

15. Asante-Owusu RA, Banham AH, Bohnert HU, Mellor EJC, Casselton LA

(1992) Heterodimerization between two classes of homeodomain proteins in the

mushroom Coprinus cinereus brings together potential DNA-binding and

transactivation domains. Gene 171: 25–31.

16. Spit A, Hyland RH, Mellor EJC, Casselton LA (1998) A role for

heterodimerization in nuclear localization of a homeodomain protein. Proc

Natl Acad Sci U S A 95: 6228–6233.

17. Banham AH, Asante-Owusu RN, Gottgens B, Thompson SAJ, Kingsnorth CS,

et al. (1995) An N-Terminal Dimerization Domain Permits Homeodomain

Proteins To Choose Compatible Partners and lnitiate Sexual Development in

the Mushroom Coprinus cinereus. Plant Cell 7: 773–783.

18. Vaillancourt LJ, Raudaskoski M, Specht CA, Raper CA (1997) Multiple genes

encoding pheromones and a pheromone receptor define the Bß1 mating-type

specificity in Schizophyllum commune. Genetics 146: 541–551.

19. Caldwell GA, Naider F, Becker JF (1995) Fungal lipopeptide mating pheromones:

a model system for the study of protein prenylation. Microbiol Rev 59: 406–422.

20. O’Shea SF, Chaure PT, Halsall JR, Olesnicky NS, Leibbrandt A, et al. (1998) A

Large Pheromone and Receptor Gene Complex determines Multiple B mating

Type Specificities in Coprinus cinereus. Genetics 148: 1081–1090.

21. Olesnicky NS, Brown AJ, Dowell SJ, Casselton LA (1999) A constitutively active

G-protein-coupled receptor causes mating self-compatibility in the mushroom

Coprinus. EMBO 18: 2756–2763.

22. Riquelme M, Challen MP, Casselton L, Brown J (2005) The origin of multiple B

mating specificities in Coprinus cinereus. Genetics 170: 1105–1119.

23. Fowler TJ, Mitton MF, Vaillancourt LJ, Raper CA (2001) Changes in

mate recognition through alterations of pheromones and receptors inthe multisexual mushroom fungus Schizophyllum commune. Genetics 158:

1491–1503.

24. James TY, Srivilai P, Kues U, Vilgalys R (2006) Evolution of the Bipolar MatingSystem of the Mushroom Coprinellus disseminatus From Its Tetrapolar Ancestors

Involves Loss of Mating-Type-Specific Pheromone Receptor Function. Genetics

172: 1877–1891.

25. Bolker M, Kahmann R (1993) Sexual pheromones and mating responses infungi. Plant Cell 5: 1461–1469.

26. Dohlman HG, Thorner J, Caron MG, Lefkowitz RJ (1991) Model systems for

the study of 7-transmembrane-segment receptors. Annu Rev Biochem 60:653–688.

27. Burnett JH (1975) Mycogenetics. London UK. John Wiley & Sons.

28. Yamazaki S, Katsuya K (1988) Mating type of pine gall rust fungus, Cronartium

quercum. Proc Jpn Acad Ser B 64: 197–200.

29. Hsueh YP, Heitman J (2008) Orchestration of sexual reproduction and virulenceby the fungal mating-type locus. Curr Opin Microbiol 11: 517–524.

30. Coelho MA, Paulo J, Goncalves P (2010) A Deviation from the Bipolar-

Tetrapolar Mating Paradigm in an Early Diverged Basidiomycete. PLoS Genet

6: e1001052. Doi:10.1371/jouornal.pgen.1001052.

31. Niculita-Hirzel H, Labbe J, Kohler A, le Tacon F, Martin F, Sanders IR, et al.

(2008) Gene organization of the mating type regions in the ectomycorrhizal

fungus Laccaria bicolor reveals distinct evolution between the two mating type loci.New Phytol 180: 329–342.

32. Ohm RA, de Jong JF, Lugones LG, Aerts A, Kothe E, et al. (2010) Genome

sequence of the model mushroom Schizophyllum commune. Nat. Biotechnol.Advanced online publication, ISSN1087-0156. pp EISSN: 1546–1696.

33. Sambrook J, Fritsch EF, Maniatis T (1989) Molecular Cloning: A Laboratory

Manual. Cold Spring Harbor NY: Cold Spring Harbor Laboratory Press.

34. Tusnady GE, Simon I (2001) The HMMTOP transmembrane topology

prediction server. Bioinformatics 17: 849–50.

35. Sonnhammer ELL, von Heijne G, Krogh A (1998) A hidden Markov model for

predicting transmembrane helices in protein sequences. In: Glasgow J,

Littlejohn T, Major F, Lathrop R, Sankoff D, Sensen C, eds. Proceedings ofthe Sixth International Conference on Intelligent Systems for Molecular Biology.

Menlo Park CA: AAAI Press. pp 175–182.

36. Finn RD, Mistry J, Tate J, Coggill P, Heger A, et al. (2010) The Pfam proteinfamilies database. Nucleic Acids Research 38 (Database issue). pp D211–D222.

doi:10.1093/nar/gkp985.

37. Lupas A, Van Dyke M, Stock J (1991) Predicting Coiled Coils from Protein

Sequences. Science 252: 1162–1164.

38. Horton P, Park K-J, Obayashi T, Nakai K (2006) Protein SubcellularLocalization Prediction with WoLF PSORT. In: Proceedings of Asian Pacific

Bioinformatics Conference APBC06 Taipei. pp 39–48.

39. Piskacek S, Gregor M, Nemethova M, Grabner M, Kovarik P, et al. PromiscuousFamily of Nine Amino Acids Transactivation Domain, 9aa TAD, Establishment

and Prediction Utilities. Department of Genetics, Max F Perutz Laboratories,Vienna Biocenter, Vienna University, Vienna 1030, Austria.

40. Piskacek S, Gregor M, Nemethova M, Grabner M, Kovarik P, et al. (2007)

Nine-amino-acid transactivation domain: establishment and prediction utilities.

Genomics 89: 756–768.

41. Niculita-Hirzel H, Hirzel AH (2008) Visualizing the gene order conservationamong genomes with ChromoMapper. International Journal of Computational

Intelligence in Bioinformatics and Systems Biology 1.

42. James TY, Liou S-R, Vilgalys R (2004) The genetic structure and diversity of theA and B mating-type genes from the tropical oyster mushroom, Pleurotus djamor.

Fungal Gen Biol 41: 813–825.

43. Thompson JD, Higgins DG, Gibson TJ (1994) CLUSTAL W: Improving thesensitivity of progressive multiple sequence alignment through sequence



weighting, positions-specific gap penalties and weight matrix choice. Nucleic

Acids Research 22: 4673–4680.44. Talavera G, Castresana J (2007) Improvement of phylogenies after removing

divergent and ambiguously aligned blocks from protein sequence alignments.

Systematic Biology 56: 564–577.45. Stamatakis A (2006) RAxML-VI-HPC: maximum likelihood-based phylogenetic

analyses with thousands of taxa and mixed models. Bioinformatics 22: 2688–2690.46. Abascal F, Zardoya R, Posada D (2005) ProtTest: Selection of best-fit models of

protein evolution. Bioinformatics 21: 2104–2105.

47. Felsenstein J (1985) Confidence limits on phylogenies: An approach using the

bootstrap. Evolution 39: 783–791.48. Gillissen B, Bergemann J, Sandmann C, Schroeer, B, Bolker M, et al. (1992) A

two-component regulatory system for self/non-self recognition in Ustilago maydis.

Cell 68: 647–657.49. Bakkeren G, Kronstad, JW (1993) Conservation of the b matingtype gene

complex among bipolar and tetrapolar smut fungi. Plant Cell 5: 123–136.50. Matheny PB, Curtis JM, Hofstetter V, Aime MC, Moncalvo J-M, et al. (2006) Major

clades of Agaricales: a multilocus phylogenetic overview. Mycologia 98: 984–997.



comparative genomics of the mating-type loci of the mushroom...

Documents